Assembly of two-dimensional (2D) layered structures into three-dimensional (3D) macroscopic hydrogel has been an enduring attracting research theme. However, the anisotropic intersheet cross-linking to form Ti3C2T x MXene-based hydrogel remains intrinsically challenging because of the superior hydrophilic nature of 2D Ti3C2T x . Herein, Ti3C2T x MXene is ingeniously assembled into the 3D macroscopic hydrogel under mild conditions by a graphene oxide (GO)-assisted self-convergence process. During the process, GO is reduced to reduced graphene oxide (RGO) by virtue of the reduction ability of Ti3C2T x , leading to the partial removal of hydrophilic oxygen-containing groups and an increase of the hydrophobicity and the π-conjugated structures of RGO, which enables the assembly of RGO into a 3D RGO framework. Simultaneously, Ti3C2T x is self-converged to be incorporated into the RGO framework by intimate interfacial interactions, thereby generating Ti3C2T x -based hydrogel. The hydrogel with interconnected porous structure holds great potential as a promising material platform for photoredox catalysis. With the incorporation of Eosin Y photosensitizer, the functional Ti3C2T x -based hydrogel exhibits enhanced photoactivity compared to the powder counterpart and features easy operability. This work enriches the rational utilization of GO/MXene colloid chemistry to design Ti3C2T x MXene-based hydrogels with improved overall efficacy in practical applications.

In session-based or sequential recommendation, it is important to consider a number of factors like long-term user engagement, multiple types of user-item interactions such as clicks, purchases etc. The current state-of-the-art supervised approaches fail to model them appropriately. Casting sequential recommendation task as a reinforcement learning (RL) problem is a promising direction. A major component of RL approaches is to train the agent through interactions with the environment. However, it is often problematic to train a recommender in an on-line fashion due to the requirement to expose users to irrelevant recommendations. As a result, learning the policy from logged implicit feedback is of vital importance, which is challenging due to the pure off-policy setting and lack of negative rewards (feedback).In this paper, we propose self-supervised reinforcement learning for sequential recommendation tasks. Our approach augments standard recommendation models with two output layers: one for selfsupervised learning and the other for RL. The RL part acts as a regularizer to drive the supervised layer focusing on specific rewards (e.g., recommending items which may lead to purchases rather than clicks) while the self-supervised layer with cross-entropy loss provides strong gradient signals for parameter updates. Based on such an approach, we propose two frameworks namely Self-Supervised Qlearning (SQN) and Self-Supervised Actor-Critic (SAC). We integrate the proposed frameworks with four state-of-the-art recommendation models. Experimental results on two real-world datasets demonstrate the effectiveness of our approach. CCS CONCEPTS• Information systems → Recommender systems; Retrieval models and ranking; Novelty in information retrieval.

No abstract

No abstract

Background: Postoperative pain management is of great importance in perioperative anesthetic care. Transversus abdominis plane (TAP) block has been described as an effective technique to reduce postoperative pain and morphine consumption after open lower abdominal operations. Meanwhile, local anesthetic infiltration (LAI) is also commonly used as a traditional method. However, the effectiveness of these two methods has not been compared before. Methods: A meta-analysis of all relevant randomized controlled trials (RCTs) was conducted to compare the efficacy of single shot TAP block with that of single shot LAI for postoperative analgesia in adults. Major medical databases and trial registries were searched for published and unpublished RCTs. The endpoints include postoperative visual analog scale (VAS) pain score, morphine requirement, and rate of postoperative nausea and vomiting (PONV). For continuous data, weighted mean differences (WMDs) were formulated; for dichotomous data, risk ratios (RR) were calculated. Results were derived using a random-/fixed-effects model with 95% confidence interval (CI).